Analysis of bronchoalveolar lavage fluid proteome from systemic sclerosis patients with or without functional, clinical and radiological signs of lung fibrosis

Available online http://arthritis-research.com/content/8/6/R160

Open AccessVol 8 No 6Research articleAnalysis of bronchoalveolar lavage fluid proteome from systemic sclerosis patients with or without functional, clinical and radiological signs of lung fibrosisAM Fietta1,4, AM Bardoni2, R Salvini2, I Passadore1, M Morosini1, L Cavagna3,4, V Codullo3,4, E Pozzi1,4, F Meloni1,4 and C Montecucco3,4

1Department of Haematological, Pneumological and Cardiovascular Sciences, University of Pavia, Via Taramelli 5, 27100 Pavia, Italy2Department of Biochemistry, University of Pavia, Via Taramelli 5, 27100 Pavia, Italy3Department of Internal Medicine, University of Pavia, Via Taramelli 5, 27100 Pavia, Italy4IRCCS San Matteo, Piazzale Golgi 2, 27100 Pavia, Italy

Corresponding author: AM Fietta, [email protected]

Received: 7 Mar 2006 Revisions requested: 21 Apr 2006 Revisions received: 31 May 2006 Accepted: 17 Oct 2006 Published: 17 Oct 2006

Arthritis Research & Therapy 2006, 8:R160 (doi:10.1186/ar2067)This article is online at: http://arthritis-research.com/content/8/6/R160© 2006 Fietta et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Lung fibrosis is a major cause of mortality and morbidity insystemic sclerosis (SSc). However, its pathogenesis still needsto be elucidated. We examined whether the alteration of certainproteins in bronchoalveolar lavage fluid (BALF) might have aprotective or a causative role in the lung fibrogenesis process.For this purpose we compared the BALF protein profileobtained from nine SSc patients with lung fibrosis (SScFib+) withthat obtained from six SSc patients without pulmonary fibrosis(SScFib-) by two-dimensional gel electrophoresis (2-DE). Onlyspots and spot-trains that were consistently expressed in adifferent way in the two study groups were taken intoconsideration. In total, 47 spots and spot-trains, correspondingto 30 previously identified proteins in human BALF, showed nosignificant variation between SScFib+ patients and SScFib-patients, whereas 24 spots showed a reproducible significantvariation in the two study groups. These latter spotscorresponded to 11 proteins or protein fragments, includingserum albumin fragments (13 spots), 5 previously recognizedproteins (7 spots), and 4 proteins (3 spots) that had not beenpreviously described in human BALF maps, namely calumenin,cytohesin-2, cystatin SN, and mitochondrial DNA

topoisomerase 1 (mtDNA TOP1). Mass analysis did notdetermine one protein-spot. The two study groups revealed asignificant difference in BALF protein composition. Whereaslevels of glutathione S-transferase P (GSTP), Cu–Zn superoxidedismutase (SOD) and cystatin SN were downregulated inSScFib+ patients compared with SScFib- patients, we observed asignificant upregulation of α1-acid glycoprotein, haptoglobin-αchain, calgranulin (Cal) B, cytohesin-2, calumenin, and mtDNATOP1 in SScFib+ patients. Some of these proteins (GSTP, Cu–Zn SOD, and cystatin SN) seem to be involved in mechanismsthat protect lungs against injury or inflammation, whereas others(Cal B, cytohesin-2, and calumenin) seem to be involved inmechanisms that drive lung fibrogenesis. Even if the 2-DEanalysis of BALF did not provide an exhaustive identification ofall BALF proteins, especially those of low molecular mass, itallows the identification of proteins that might have a role in lungfibrogenesis. Further longitudinal studies on larger cohorts ofpatients will be necessary to assess their usefulness aspredictive markers of disease.

IntroductionThe analysis of the bronchoalveolar lavage fluid (BALF) pro-teome can potentially provide important information about

changes in protein expression and secretion during the courseof pulmonary disorders. At the moment, more than 100 humanproteins or protein isoforms have been identified in human

Page 1 of 11(page number not for citation purposes)

Anti-Scl70 = anti-topoisomerase I antibodies; BAL = bronchoalveolar lavage; BALF = bronchoalveolar lavage fluid; Cal = calgranulin; 2-DE = two-dimensional gel electrophoresis; GSTP = glutathione S-transferase P; HRCT = high-resolution computed tomography; IL = interleukin; IPG = immo-bilized pH gradient; MAP = mitogen-activated protein; MS, mass spectrometry; mtDNA TOP1 = mitochondrial DNA topoisomerase 1; SOD = super-oxide dismutase; SSc = systemic sclerosis; SScFib- = systemic sclerosis patients without lung fibrosis; SScFib+ = systemic sclerosis patients with lung fibrosis; TLCO = carbon monoxide transfer.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Abstract&list_uids=17044913

http://arthritis-research.com/content/8/6/R160

http://creativecommons.org/licenses/by/2.0

http://www.biomedcentral.com/info/about/charter/

Arthritis Research & Therapy Vol 8 No 6 Fietta et al.

BALF by using two-dimensional gel electrophoresis (2-DE),including both proteins released locally in the lung by inflam-matory or epithelial cells and proteins derived from serum bydiffusion across the capillary–alveolar barrier [1,2]. Previousstudies showed that the protein composition of BALF wasaltered in sarcoidosis, idiopathic pulmonary fibrosis, allergicasthma, and chronic obstructive pulmonary disease [3-5]. Fur-thermore, Rottoli and colleagues [6] recently demonstratedthat systemic sclerosis patients with pulmonary fibrosis (SSc-

Fib+) showed a BALF profile that was intermediate betweenthat of patients with sarcoidosis and idiopatic pulmonary fibro-sis, despite some particularities.

Systemic sclerosis (SSc) is an autoimmune disorder ofunknown origin, characterized by an excessive deposition ofcollagen and other extracellular matrix proteins on the skin aswell as multiple internal organs, alterations in microvascula-ture, and humoral and cellular abnormalities [7]. Clinically, thedisease is very heterogeneous, ranging from limited skininvolvement with limited systemic alterations to clinical pic-tures with diffuse skin sclerosis and severe visceral involve-ment with fibrosis, microvascular abnormalities andmononuclear cell infiltration of the gastrointestinal tract, lungs,heart, and kidneys. Pulmonary involvement has emerged aspotentially the most serious visceral lesion in SSc. Alveolitis ismore commonly seen in SSc with diffuse skin sclerosis, partic-ularly in the presence of serum anti-topoisomerase I (anti-Scl70) antibodies, but only 15% of patients eventually reachsevere end-stage lung fibrosis [8].

Given the complexity of the pathogenesis of lung fibrosis andthe lack of biological markers that can reliably predict whichpatients will evolve the disease, great interest has been shownin these issues, also comparing cohorts of SSc patients withor without lung involvement. The ability to identify the proteinsthat are differently expressed in the lower airways of SScFib+patients from those in SScFib- patients might provide newinsight into the disease pathogenesis and into the identifica-tion of possible protein biomarkers of the disease's progres-sion. With this aim, we used the 2-DE analysis to assess andcompare the protein profile of BALFs from a group of SScFib+and SScFib- patients, respectively.

Materials and methodsPatientsThe samples consisted of archived BALF samples acquiredbetween March 2004 and April 2005 from 15 SSc patientswho satisfied the preliminary American College of Rheumatol-ogy criteria for disease classification [9]. All the patients werefemale and non-smoker outpatients at the Rheumatology Unitof the Policlinico San Matteo, Pavia, Italy. They all gave theirinformed consent to undergo bronchoscopy, and none werebeing treated with prednisone or other immunosuppressiveagents at the time of enrollment. Five patients had SSc withlimited skin involvement and 10 had SSc with diffuse skin scle-

rosis; indirect immunofluorescence on Hep-2 cells detectedthe antinuclear antibodies, and the anti-Scl70 antibodies wereevaluated by enzyme-linked immunosorbent assay (DiamedixCo., Miami, FL, USA). All the patients tested positive for anti-nuclear antibodies, and nine had anti-Scl70 antibodies.

The diagnosis of SSc-associated lung fibrosis was based onclinical, functional and radiographic signs. In all the cases, car-bon monoxide transfer, forced vital capacity, chest radiogra-phy, and high-resolution computed tomography (HRCT)determined the respiratory involvement, which was defined asat least a grade 1 severity on the disease severity scale forSSc [10]. In all the cases, the same expert radiologist calcu-lated the Kazerooni score for fibrosis as described previously[11]. Nine patients showed a decrease of at least 20% withregard to the expected carbon monoxide transfer or forcedvital capacity values [12]. This decrease was associated withabnormal chest radiograph and the presence of significantfibrosis on the HRCT scan (mean Kazerooni fibrosis score: 10± 5 SD). These nine patients were included in the SScFib+group. The remaining six patients, who revealed no evidenceof lung involvement either on the lung function tests or on theHRCT scan (absence of fibrosis and/or a ground-glass pat-tern), were included in the SScFib- group. Table 1 details thepatients' demographic, clinical, immunological and functionalcharacteristics.

Bronchoalveolar lavage and phenotype analysisBronchoalveolar lavage (BAL) was performed as describedpreviously [13]. To summarize briefly, the distal tip of the bron-choscope was wedged into the middle lobe or lingular bron-chus. A total of 150 ml of warm sterile saline solution wereinstilled in 30 ml aliquots and serially recovered by gentle aspi-ration. The first collected aliquot was used for the microscopicand cultural examination of common bacteria and fungi, and ofdirect acid-fast bacilli smears (Kinyoun method) and culturesfor mycobacteria, and for the microscopic examination ofPneumocystis jiroveci (Gomori-Grocott silver stain). All theother aliquots were used to assess the total and differentialcell counts (performed on cyto-centrifuged preparations withthe use of May–Grünwald–Giemsa and Papanicolaou stain-ings) and the percentages of CD4+ and CD8+ T cells(assessed by cytofluorimetry). Samples were immediately fil-tered twice through gauze, then centrifuged at 1,500 r.p.m. for10 min, and BALF samples were promptly frozen in aliquotsand stored at -80°C until used.

Two-dimensional gel electrophoresisBALF samples were dialyzed (tubing cut-off 1 kDa) for a fewdays against several changes of distilled water, in the pres-ence of protease inhibitors (protease inhibitor cocktail; Sigma,St Louis, MO, USA), and the protein content was assessed(bicinchoninic acid protein assay kit; Pierce, Rockford, IL,USA). Samples were freeze-dried and pellets were suspendedin 8 M urea, 4% w/v CHAPS, 65 mM 1,4-dithioerythritol (solu-



tion A). Two-dimensional gel electrophoresis was performedas described previously [14]. In brief, for each analytical exper-iment the same amount (70 μg) of proteins was loaded onimmobilized pH gradient (IPG) strips (pH 3.5 to 10, 18 cm;Amersham Biosciences, Uppsala, Sweden), which were swol-len in solution A plus 0.8% carrier ampholytes (Ampholine™,pH 3.5 to 10 for IEF; Amersham Biosciences) and a trace ofbromophenol blue. A larger amount of protein (1 mg) had to beloaded for identification by mass spectrometry (MS). Rehydra-tion was performed in the strip holder of the IPGphor systemat 20°C (Amersham Biosciences) and isoelectrofocusing wasterminated at 60 kV h. The IPG strips were equilibrated first for12 minutes in urea/SDS/Tris buffer, and then in the samebuffer containing 2.5% (w/v) iodoacetamide. The run in thesecond dimension was performed on 9 to 16.5% polyacryla-mide linear gradient gels with a constant current of 40 mA pergel at 10°C. The gels were stained with ammoniacal silvernitrate or colloidal Coomassie G-250 [15].

Evaluation and protein identificationThe Versadoc Imaging System Model 3000 (Bio-Rad, Rich-mond, CA, USA) scanned the gels, which the PD QUEST 7.1software (Bio-Rad) then analyzed. The total number of spots,the non-matched spots and the spot parameters were meas-ured. To detect differentially expressed protein spots in SSc-

Fib+ compared with those in SScFib- patients, the software wasinstructed to create one reference map for each patient group.The software established these maps by taking into accountonly the spots that were constantly expressed in each patient

of the same group. At least two gels from each patient wereevaluated. Initially, we compared the two reference maps so asto assess differences in spot quantity between the two studygroups. The PD QUEST 7.1 software determined the spotquantity by formula: Spot height × π × σx × σy (where the spotheight represented the peak of the spot's Gaussian represen-tation, and σx and σy were the SD of the spot's Gaussian dis-tribution in the directions of the x and y axes, respectively)normalized the value for the total protein content of each gel.The differences between the two study groups were then con-firmed by analysing the spot quantities observed in singlepatients. On the basis of the recent guidelines for proteomicsresearch [16], non-matched spots and spots that differed sig-nificantly in quantity (p < 0.05) were regarded as differentiallyexpressed in SScFib+ compared with SScFib- patients.

Proteins were identified by comparing the BALF mapsobtained in this study with the SWISS-2D PAGE humanplasma map [17] and published BALF maps [5,6], usingimmunoblotting, matrix-assisted laser desorption/ionizationMS, and/or liquid chromatography MS/MS. For the MS analy-sis, the selected spots, excised from Coomassie-stained gels,were analyzed in a Voyager DE-PRO mass spectrometer(Applied Biosystems, Foster City, CA, USA), equipped with areflectron analyzer and used in delayed extraction mode. Masssignals were used for database searching with the MASCOTpeptide fingerprinting search program (Matrix Science, Bos-ton, MA, USA). A quadrupole time-of-flight. Ultima hybrid massspectrometer (Micromass, Toronto, Ontario, Canada) was

Table 1

Demographic, clinical, immunological and functional characteristics of scleroderma patients

Characteristic SScFib+ SScFib-

Number 9 6

Age, years (mean ± SD) 55.6 ± 10.3 50.5 ± 9.6

Cutaneous involvement

Lc SSc (percentage) 1 (11.1) 4 (66.7)

Dc SSc (percentage) 8 (89.9) 2 (33.3)

Skin score (mean ± SD) 17.0 ± 8.0 10.0 ± 6.0

Autoantibodies

ANA positive (percentage) 9 (100) 6 (100)

Anti-Scl70 positive (percentage) 7 (77.7) 2 (33)

Lung function test

FVC, percentage predicted (mean ± SD) 80.9 ± 11.7a 118.7 ± 11.6

TLCO, percentage predicted (mean ± SD) 60.01 ± 12.6a 72.03 ± 13.8

Kazerooni score 10.0 ± 5 0

Skin score and Kazerooni score for fibrosis were determined as described previously [9,11]. SScFib+, systemic sclerosis patients with functional lung fibrosis and signs of lung fibrosis on high-resolution computed tomography; SScFib-, SSc patients with no signs and symptoms of lung involvement; Lc SSc, limited cutaneous SSc; Dc SSc, diffuse cutaneous SSc; ANA, antinuclear antibodies; anti-Scl70, anti-topoisomerase I antibodies; FVC, forced vital capacity; TLCO, carbon monoxide transfer. aDecrease more than 20% with regard to the expected TLCO or FVC values.



used for the liquid chromatography–MS/MS analysis. Theinstrument was set up in a data-dependent MS/MS mode, andthe raw MS and MS/MS spectra were elaborated with Protein-Lynx software. The MASCOT MS/MS ion search software forprotein identification was used. These evaluations were per-formed at CEINGE Biotecnologie Avanzate s.c. a r.l., Univer-sity Federico II, Naples, Italy).

Statistical methodsResults are presented as mean ± SD or median and interquar-tile range. Normally distributed variables were analyzed by atwo-tailed Student's t test. Proteomics expression data werecompared by using the Mann–Whitney U test. For all tests p< 0.05 was regarded as significant. Analyses were performedwith PRISM 3.0 (GraphPad Software, San Diego, CA, USA).

ResultsBAL was performed in 15 SSc patients, including 9 patientswith clinical, functional and HRCT signs of interstitial lungfibrosis (SScFib+ patients) and 6 patients who had no signsand symptoms of lung involvement (SScFib- patients). No clini-cal or microbiological evidence of bacterial or fungal infectionswas ever found in the included patients. Furthermore, the twopatient groups revealed no significant difference in total BALcells recovered, CD4+/CD8+ T-cell ratio, protein content, ordifferential BAL cell counts, although the SScFib+ patients dis-played a trend toward higher neutrophil and eosinophil counts(Table 2). Neutrophil and eosinophil counts, as expected, weresignificantly higher in the SScFib+ patients than in dataobtained in our laboratory from a group of healthy volunteers[12].

BALF proteome in SScFib+ and SScFib- patientsTwo-dimensional gel electrophoresis analyzed the whole pro-tein profile of BALF from SScFib+ and SScFib- patients, to eval-uate possible differences in protein expression between thetwo states of the disease. The mean number of spots, of whichseveral were organized as spot-trains, in gels of BALF from

SScFib+ and SScFib- patients was 365 ± 84 and 390 ± 75,respectively. However, only 332 and 348 of these spots wereconstantly expressed in BALF from all single SScFib+ and SSc-

Fib- patients, respectively. Figure 1 shows a representativeCoomassie-stained gel map from an SScFib+ patient (Figure1a) and an SScFib- patient (Figure 1b). By comparing the quan-tity of 59 individual spots and 12 spot-trains, we identified 35spots and 12 spot-trains whose quantity did not vary signifi-cantly between SScFib+ and SScFib- patients (Mann–WhitneyU test p > 0.05). By matching these gel maps with publishedtwo-dimensional BALF maps, we identified 30 proteins corre-sponding to the 47 spots and spot-trains that did not vary inquantity between the two patient groups (Table 3). Immunob-lotting also validated the assignment of some of these pro-teins. Furthermore, we detected 24 individual spots thatshowed a reproducible, significant variation between the twostudy groups. The quantity in 21 of these spots was signifi-cantly increased, whereas the quantity in 3 spots was signifi-cantly decreased in SScFib+ patients compared with SScFib-patients.

Matrix-assisted laser desorption/ionization MS with or withoutliquid chromatography–MS/MS analysis identified the pro-teins corresponding to 21 spots that were constantly and sig-nificantly expressed in a different way between the two studygroups. A comparison with published two-dimensional BALFmaps followed by immunoblotting validation identified the pro-teins corresponding to the remaining three spots with theaforementioned criteria (Table 4). These 24 differentlyexpressed spots corresponded to 11 proteins/protein frag-ments in total: 13 spots corresponded to serum albumin frag-ments, 7 of them corresponded to 5 previously recognizedBALF proteins (α1-acid glycoprotein (2 spots), haptoglobin-αchain (2 spots), calgranulin (Cal) B, Cu–Zn superoxide dis-mutase (SOD) and glutathione S-transferase P (GSTP)), and3 spots corresponded to 4 proteins that, to our knowledge,had not previously been described in published BALF two-dimensional maps (calumenin and cytohesin-2 (1 spot), cysta-

Table 2

Cytology, CD4+/CD8+ T-cell ratio and protein contents of bronchoalveolar lavage samples

Parameter SScFib+ (n = 9) SScFib- (n = 6) Controls (n = 9)

Total BAL cells (105/ml) 2.05 ± 1.13 1.80 ± 0.74 2.02 ± 1.33

Macrophages (percentage) 84.97 ± 8.66 87.73 ± 5.12 85.78 ± 7.09

Lymphocytes (percentage) 8.01 ± 4.15 7.80 ± 2.63 12.33 ± 6.36

Neutrophils (percentage) 4.44 ± 4.03a 2.76 ± 1.98 1.56 ± 1.59

Eosinophils (percentage) 2.62 ± 3.66a 0.83 ± 0.60 0.33 ± 0.70

CD4+/CD8+ ratio 2.30 ± 1.02 2.14 ± 1.16 2.4 ± 0.71

Protein concentration (μg/ml) 130.44 ± 53.76 133.43 ± 32.38 ND

Results are given as mean ± SD. SScFib+, systemic sclerosis patients with functional lung fibrosis and signs of lung fibrosis on high-resolution computed tomography; SScFib-, SSc patients with no signs and symptoms of lung fibrosis. Control values are taken from data previously assessed in a group of healthy volunteers [12]. BAL, bronchoalveolar lavage; ND, not determined. ap < 0.05 compared with controls (all assessed by Student's t test).



tin SN and a fragment of mtDNA TOP1). One protein spot wasmarked as not determined because its identification gave ascore that indicated uncertain identification. SScFib+ patientsshowed a significant upregulation of α1-acid glycoprotein,haptoglobin-α chain, Cal B, calumenin, and cytohesin-2 levelscompared with SScFib- patients. Furthermore, the expressionof a fragment of mtDNA TOP1 was exclusively detectable inSScFib+ patients, including two patients who were negative foranti-Scl70 serum antibodies. Finally, whereas Cu–Zn SODwas significantly downregulated, GSTP and cystatin SN werenot detectable in SScFib+ patients in contrast with SScFib-patients.

DiscussionThe pathogenesis of SSc is extremely complex, and no singleunifying hypothesis or pathogenetic mechanism can explain allthe aspects of this disease. It is therefore widely accepted thatfunctional alterations of several cell effectors, including fibrob-lasts, endothelial cells, and T and B lymphocytes, are intimatelyinvolved in the development of the clinical and pathologicalmanifestations of SSc [7,18,19]. Abnormalities in lung func-tion occur in about 70% of SSc patients and currently repre-sent the main cause of mortality in this population [7].However, the factors that drive the severe and progressive vis-ceral fibrosis remain mostly undetermined. Unlike the recentpathogenetic hypothesis for idiopathic interstitial lung fibrosis,which assigns only a marginal role to inflammation, fibrogene-sis in SSc-related lung involvement is generally thought to bethe result of a complex series of interstitial immunoinflamma-tory reactions [20]. The persistent and unregulated activationof genes that encode various collagen and other extracellularmatrix proteins is regarded as a crucial pathogenetic event[21,22]. Furthermore, an alteration in the production of pro-inflammatory and/or pro-fibrotic cytokines and chemokines,growth factors (in particular transforming growth factor-β), andsignaling molecules also seems to have a relevant role in fibro-genesis [23,24]. The identification of proteins that are differ-ently expressed in the airways of SScFib+ patients comparedwith patients with no signs or symptoms of lung involvementmay cast light on the pathogenetic mechanism of the diseaseand may thereby indicate protein biomarkers that are useful forthe diagnosis of lung involvement, for the monitoring of SScprogression, and for the identification of possible new thera-peutic targets. In this context, the use of 2-DE and the map-ping of BALF seem highly interesting.

Our results demonstrated that quantitative and qualitative dif-ferences can be found between patients with SSc-associatedlung fibrosis and patients with SSc with no signs or symptomsof lung involvement. At least 24 well-defined spots showedsignificant and reproducible variations in relative abundancebetween the two patient groups. Some of these spots corre-sponded to previously described BALF proteins and proteinfragments, namely α1-acid glycoprotein, GSTP, Cu–Zn SOD,haptoglobin-α chain, and calgranulin B, whereas 4 of the pro-

teins found had not previously been described in publishedtwo-dimensional BALF maps, namely cytohesin-2, calumenin,cystatin SN, and human mtDNA TOP1 fragment. In SScFib+patients, GSTP and cystatin SN were below the detectionlimit, whereas Cu–Zn SOD was downregulated. In contrast,α1-acid glycoprotein, calumenin, cytohesin-2, haptoglobin-αchain, and calgranulin B were significantly upregulated, andthe expression of mtDNA TOP1 was observed only in SScFib+patients.

On the basis of these preliminary results, we can infer that 2-DE analysis of BALF can provide useful insight into the under-standing of mechanisms involved in lung fibrogenesis throughthe identification of proteins that are deregulated in SSc-asso-ciated lung fibrosis. However, our results, as well as those ofprevious studies, clearly indicated that 2-DE analysis did notprovide an exhaustive identification of all proteins that can bepresent in BALF, especially those of low molecular mass suchas cytokines, chemokines, growth factors, and signaling mole-cules. Because these molecules are thought to be relevant tothe pathogenesis of SSc-associated lung fibrosis, a combina-tion of 2-DE analysis and proteomic approaches that specifi-cally address the identification of these factors is necessary forthe assessment of the whole protein pattern of BALF. Despitethese considerations, most of the proteins we found differentlyregulated between SScFib+ patients and SScFib- patients couldhave a role in the development of lung fibrosis, whereby someof them are involved in the mechanisms that protect againstlung injury or inflammation whereas others are involved in themechanisms that drive fibroproliferation.

Oxidative stress is believed to be important in the pathogene-sis of lung damage and in the development of lung fibrosis[25]. Among various enzymatic and non-enzymatic mecha-nisms that protect cells and tissues from oxidants [26,27], glu-tathione transferases [28] and SODs [29] are among suchmechanisms that are thought to have a key protective role,especially in the lung. High levels of extracellular SODs arethought to have a protective role for lung matrix [29]. Recently,Tourkina and colleagues [30] demonstrated a deficiency inGSTP1 in scleroderma lung fibroblasts. The observation thatGSTP and Cu–Zn SOD levels are severely downregulated inBALF from SScFib+ patients is in line with these findings, andconfirms the relevance of these molecules as a defense toolfor the lung against oxidant-induced damage and/or fibropro-liferation.

Moreover, in BALF from SScFib+ patients, we did not detectlevels of cystatin SN, which was, in contrast, well representedin SScFib- patients. Cystatin SN is a secreted protein, whichbelongs to SD-type or type-2 cystatins [31], which are potentinhibitors of CA clan mammalian cysteine peptidases. Intracel-lularly, these enzymes participate in normal protein turnover, inantigen and protein processing, and in apoptosis, whereasextracellularly they are involved in the inhibition of the proteo-




Figure 1

Representative Coomassie-stained gels of bronchoalveolar lavage fluidRepresentative Coomassie-stained gels of bronchoalveolar lavage fluid. Samples shown were taken from a systemic sclerosis patient with lung fibro-sis (SScFib+) (a) and a systemic sclerosis patient without lung fibrosis (SScFib-) (b). Analyzed spots are marked with a number and the corresponding identified proteins are listed in Tables 3 and 4. Proteins whose expression was significantly different between SScFib+ patients and SScFib- patients are also indicated by name.


lytic cascade [32]. The direct inhibition of endogenous andexogenous cysteine peptidases is the only function describedso far for cystatin SN, suggesting that its function is to protecttissues against injury induced by these enzymes. However,recent results obtained with other type-2 cystatins seem tosuggest a broader spectrum of activities, including the upreg-

ulation of IL-6 by human gingival fibroblasts, interferon-γexpression in CD4+ T cells [33,34], and the involvement in thedifferentiation and maturation of human dendritic cells [35].

Furthermore, we found an upregulation of at least six proteinsin BALF from SScFib+ patients: haptoglobin-α chain, α1-acid

Table 3

Proteins whose quantity in BALF did not significantly vary between SScFib+ and SScFib- patients

Spot no. Protein Identification method

1 1β glycoprotein GM

2 Hemopexin GM

3 Transferrin GM

4 Albumin GM

5 Ig heavy chain α GM

6 Vitamin D-binding protein GM

7 α1-antitrypsin GM, Ab

8 α1-antichymotrypsin GM, Ab

9 α1-HS-glycoprotein GM

10 Leucin-rich α2-glycoprotein GM

11 Fibrinogen γ A chain GM, Ab

12 Ig heavy chain γ GM

14 Haptoglobulin 1β chain GM

15 Pulmonary surfactant-associated protein A GM

16 Apolipoprotein A1 GM, Ab

17 Ig light chain κ, λ GM, Ab

18 Serum retinol-binding protein GM

20 Transthyretin GM, Ab

21 Immunoglobulin-binding factor GM

22 Calgranulin A GM

23 Hemoglobin β chain GM

24 ERP60 (protein disulfide isomerase) GM, Ab

25 α1-antitrypsin (fragment) GM, Ab

26 Ig J chain GM

27 C-reactive protein GM

28–30 Apolipoprotein A1 (fragment) GM, Ab

31 Calcyphosine GM

32 TCTP (translationally controlled tumor protein) GM

33 ATP synthase δ chain mitochondrial GM

34 L-FABP (liver fatty acid-binding protein) GM

Spot quantity was assessed by PDQUEST 7.1 software on two BALF gel maps of individual SScFib+ (n = 9) and SScFib- (n = 6) patients. Spot quantity between SScFib+ and SScFib- patients did not differ significantly (Mann–Whitney U test, p > 0.05). Spot no. refers to the annotations in Figure 1. BALF, bronchoalveolar lavage fluid; SScFib+, systemic sclerosis patients with functional lung fibrosis and signs of lung fibrosis on high-resolution computed tomography; SScFib-, SSc patients with no signs and symptoms of lung fibrosis; GM, gel matching with plasma two-dimensional gel electrophoresis database from Swiss-PROT [17] and published two-dimensional gel electrophoresis BALF maps [5,6]; Ab, immunoblotting.



glycoprotein, calumenin, cytohesin-2, Cal B, and mtDNATOP1.

Haptoglobin and α1-acid glycoprotein are positive acute-phase proteins, whose plasma concentrations increase duringinflammation [36]. The increased levels of both proteins inBALF from SScFib+ patients is likely to originate from the pas-sive diffusion from plasma, as a result of the increased perme-ability of the air–blood barrier during the course of fibrosingalveolitis.

Calumenin is a member of the CREC family [37] that issecreted into the medium of cultured cells including fibrob-lasts [38] and early melanosomes [39]. Intracellularly, calu-menin has a chaperone function in several Ca2+-regulatedprocesses, but extracellularly its function is not yet fully under-stood. Vorum and colleagues found that calumenin binds tothe serum amyloid P component, suggesting that it is involvedin the formation of amyloid deposits in different tissues [40].Coppinger and colleagues recently described the presence ofcalumenin in platelets and its release on platelet activation, aswell as the protein's presence in human atheroscleroticlesions [41], therefore suggesting a possible role of this pro-tein in leading to vascular injury through the promotion of plate-

let adhesion, fibrinogenesis, and intravascular thrombusdeposition. Taking into account that a crucial pathogeneticstep in SSc involves injury to microvasculature [19,20], it ispossible to infer that calumenin could have a role in takingsuch an injury to lung level.

Cytohesin-2 is a member of a family of cytoplasmic signalingproteins [42], which may move from the cytosol to the plasmamembrane after cell stimulation [43]. Although its biologicalfunction is mostly unknown, a recent study demonstrated thatcytohesin-2 is an activator of the mitogen-activated protein(MAP) kinase signaling pathway [44]. Because MAP kinasesare activator signals for lung fibroblasts [23], we suggest thatcytohesin-2 might participate in the development and/or pro-gression of lung fibrosis in SSc patients through the activationof the MAP kinase signaling pathways in lung fibroblasts. Thishypothesis needs to be verified with further experimentalobservations.

Calgranulin B is an S100 protein that is unique together withCal A in its myeloid-specific expression profile and in its abun-dance in neutrophils [45]. Increased levels of Cal B arepresent in the bronchial secretion of patients with chronicinflammatory disease, and these levels may induce the produc-

Table 4

BALF proteins whose regulation in SScFib+ and SScFib- patients is different

Spot no. Median (IQR) spot quantity Protein or fragment MW (kDa) pI AC Identification method

SScFib+ SScFib- p

13 328.0 (256.5–482.5) 86.0 (68.6–110.4) 0.0003 α1-acid glycoprotein (orosomucoid) 23.5 4.93 P02763 GM (plasma, BALF), Ab

19 7.9 (5.6–9.6) 28.2 (24.5–32.6) Cu–Zn superoxide dismutase 15.8 5.70 P00441 GM (plasma, BALF), Ab

35 343.0 (328.5–664.0) 168.0 (131.5–267.0) 0.0041 Serum albumina 69.4 5.92 P02768 MS

36 96.6 (68.9–127.6) 24.4 (17.2–64.5) 0.0012 Calgranulin B (S100A9; MRP14) 13.2 5.71 P06702 MS, GM (plasma, BALF)

37 53.3 (47.4–72.1) 11.1 (8.0–22.9) 0.0001 Haptoglobin (Hp2) α chaina 15.9 5.57 P00738 MS, LC, GM (plasma, BALF)

38 Not expressed 32.0 (29.5–39.0) Glutathione S transferase P 23.2 5.44 P09211 MS, GM (plasma, BALF)

39 59.1 (50.8–73.1) 14.8 (9.4–19.6) 0.0014 Cytohesin-2a,b 46.5 5.43 Q99418 MS, LC MS, LC

Calumenina,b 37.1 4.47 O43852 MS, LC MS, LC

40 29.2 (25.5–30.8) Not expressed Mitochondrial DNA topoisomerase 1a,b 69.8 9.46 Q969P6 MS, LC

41 Not expressed 45.5 (32.2–58.9) Cystatin SNb 16.3 6.82 P01037 MS, LC

42 Not expressed 37.7 (34.0–41.0) Not determined - - - (MS, LC)

Spot quantity values were measured on two BALF gel maps of individual SScFib+ (n = 9) and SScFib- (n = 6) patients by PDQUEST 7.1 software. Results are expressed as median and interquartile range (IQR). Spot no. refers to the annotations in Figure 1; p values were calculated with the Mann–Whitney U test. BALF, bronchoalveolar lavage fluid; SScFib+, systemic sclerosis patients with functional lung fibrosis and signs of lung fibrosis on high-resolution computed tomography; SScFib-, SSc patients with no signs and symptoms of lung fibrosis; MW, theoretical average molecular mass; pI, isoelectric point; AC, accession number from the Swiss-PROT 2DPAGE database [17]; GM, gel matching with plasma two-dimensional gel electrophoresis database from Swiss-PROT [17] and published BALF two-dimensional gel electrophoresis maps [5,6]; Ab, immunoblotting; MS, matrix-assisted laser desorption/ionization mass spectrometry analysis; LC, liquid chromatography–tandem mass spectrometry analysis. aFragments. bProteins not previously described in published BALF two-dimensional gel electrophoresis maps.



tion of IL-8 by airway epithelial cells [46]. Furthermore, Vie-mann and colleagues demonstrated that the expression of CalB and Cal A correlates with the inflammatory activity in sys-temic vasculitis, thereby confirming the role of these proteinsin inflammatory reactions of endothelia [47]. Although the gen-eral consensus is that the Cal B function depends mainly onheterodimer formation (Cal B–Cal A), several studies havedemonstrated that the monomer is also a potent stimulator ofneutrophils and is involved in the recruitment of leukocytes tothe site of inflammation through the regulation of adhesion andthe extravasion of neutrophils [45,48]. Furthermore, recom-binant Cal B and its homodimer stimulate the proliferation ofrat fibroblasts, suggesting that this protein might also functionas a mitogen for fibroblasts in chronic inflammation [49]. Onthe basis of these observations, Cal B could have a role inSSc-associated lung fibrosis through several pathways,including amplifying the inflammatory process by inducing theextravasion of neutrophils and the production of IL-8, inducingan inflammatory reaction in the endothelia, and/or stimulatinglung fibroblast proliferation.

Finally, we found that a fragment of mtDNA TOP1 was detect-able only in BALF from patients with interstitial lung fibrosis,

including in two SScFib+ patients who were negative for serumanti-Scl70 antibodies. This protein fragment was neverobserved in BALF from SScFib- patients, including in twopatients with positive anti-Scl70 antibodies. A common 13-exon motif and a similar amino-acid sequence characterizegenes for mitochondrial and nuclear DNA TOP1, except forthe N-terminal domain, which in mtDNA TOP1 contains a mito-chondrial localization sequence instead of the nuclear localiza-tion signal [50]. Although we demonstrated that the presenceof detectable levels of mtDNA TOP1 was a constant and par-ticular feature of SScFib+ patients, we did not fully ascertain theantigenic role of this protein fragment and its relation with thedevelopment of lung fibrosis. Because autoantibodies againstDNA TOP1 are strongly correlated with interstitial pulmonaryinvolvement in SSc [51], further studies are warranted toassess the role and the frequency of this antigenic target inBALF.

ConclusionOn the basis of these preliminary results we believe that thecomparative analysis of BALF proteome from SScFib+ andSScFib- patients by 2-DE might provide a useful insight intosome relevant steps of lung fibrogenesis. SSc-associated lung

Figure 2

Hypothetical role of some disregulated BALf proteins in mechanisms driving SSc-related lung fibrosisHypothetical role of some disregulated BALf proteins in mechanisms driving SSc-related lung fibrosis. Inflammation is thought to be the main mech-anism driving lung fibrosis in scleroderma patients. In this complex inflammatory process several pathways are involved, including the activation of T cells and epithelial cells, the secretion of pro-inflammatory cytokines and growth factors, and fibroblast proliferation. Furthermore, products from the coagulation cascade and oxidative stress may contribute to fibrogenesis. The upregulation of calgranulin B (Cal B), cytohesin-2 and calumenin might favor inflammation and fibrogenesis, whereas the downmodulation of the protective factors glutathione S-transferase P (GSTP), Cu–Zn superoxide dismutase (SOD) and cystatin SN may amplify tissue injury and inflammation. MMP, matrix metalloproteinases; TIMP, tissue inhibitor of matrix metal-loproteinases; ROS, reactive oxygen species; RNS, reactive nitrogen species; O2

-, superoxide; H2O2, hydrogen peroxide; NO, nitric oxide; SScFib+, systemic sclerosis patients with lung fibrosis; SScFib-, systemic sclerosis patients without lung fibrosis.



fibrosis is a multifaceted process involving the activation of Tcells and epithelial cells, the secretion of pro-inflammatorycytokines and growth factors, and fibroproliferation (Figure 2).In addition, several factors, including activation of the coagula-tion cascade and oxidative stress, are thought to influence thedevelopment of lung fibrosis. At least six proteins that were dif-ferently regulated in BALF from SScFib+ patients and in BALFfrom SScFib- patients might have a role in the pathogenesis ofSSc-associated interstitial lung disease. As shown in Figure 2,the upregulation of Cal B, cytohesin-2, and calumenin in BALFfrom SScFib+ patients could further promote the synthesis ofpro-inflammatory cytokines, the influx of inflammatory cells,and fibrinogenesis, whereas the downregulation of GSTP,Cu–Zn SOD and cystatin SN could favor vascular and lung tis-sue injury through the impairment of mechanisms that protectagainst oxidants and cysteine peptidases. However, furtheranimal and human studies will be necessary both to assess theactual role of each of these factors in the development of SSc-associated lung fibrosis and to determine whether these fac-tors are relevant as disease biomarkers, and if so which ones.Conclusions from these studies could enable us to predictwhich patients will develop progressive pulmonary fibrosis.

Competing interestsThe authors declare that they have no competing interests.

Authors' contributionsFAM generated and planned the project, supervised the workand wrote the manuscript. BAM and SR contributed equally toperform the research, analyzed critically data and participatedin the preparation of the manuscript. PI assisted in designingthe experiments and conducted 2-DE and immunoblotting.MM was involved in data analysis and performed statisticalanalysis. CL and CV selected patients and revised all clinicalcharts. PE was involved in the coordination of the study and inthe discussion and revision of the manuscript. MF participatedin the study design, supervised the work, read the bronchoal-veolar lavage and helped to draft the manuscript. MC partici-pated in the study design and its coordination and participatedin preparation of the manuscript. All authors read andapproved the final manuscript.

AcknowledgementsWe thank Dr Angela Flagiello for help with mass analysis, and Professor Piero Pucci for insightful comments on the manuscript. This work was supported by Fondazione Cariplo, Italy (Grant Rif.2003.1644/10.8485).

References1. Wattiez R, Hermans C, Bernard A, Lesur O, Falmagne P: Human

bronchoalveolar lavage fluid: two-dimensional gel electro-phoresis, amino acid microsequencing and identification ofmajor proteins. Electrophoresis 1999, 20:1634-1645.

2. Wattiez R, Hermans C, Cruyt C, Bernard A, Falmagne P: Humanbronchoalveolar lavage fluid protein two-dimensional data-base: study of interstitial lung diseases. Electrophoresis 2000,21:2703-2712.

3. Magi B, Bini L, Perari MG, Fossi A, Sanchez JC, Hochstrasser D,Paesano S, Raggiaschi R, Santucci A, Pallini V, et al.: Bronchoal-veolar lavage fluid protein composition in patients with sar-

coidosis and idiopathic pulmonary fibrosis: a two-dimensionalelectrophoretic study. Electrophoresis 2002, 23:3434-3444.

4. Sabounchi-Schutt F, Astrom J, Hellman U, Eklund A, Grunewald J:Changes in bronchoalveolar lavage fluid proteins in sarcoido-sis: a proteomics approach. Eur Respir J 2003, 21:414-420.

5. Wattiez R, Falmagne P: Proteomics of bronchoalveolar lavagefluid. J Chromatogr B Analyt Technol Biomed Life Sci 2005,815:169-178.

6. Rottoli P, Magi B, Perari MG, Liberatori S, Nikiforakis N, BargagliE, Cianti R, Bini L, Pallini V: Cytokine profile and proteome anal-ysis in bronchoalveolar lavage of patients with sarcoidosis,pulmonary fibrosis associated with systemic sclerosis and idi-opathic pulmonary fibrosis. Proteomics 2005, 5:1423-1430.

7. Jimenez SA, Derk CT: Following the molecular pathwaystoward an understanding of the pathogenesis of systemicsclerosis. Ann Intern Med 2004, 140:37-50.

8. Basu D, Reveille JD: Anti-scl-70. Autoimmunity 2005, 38:65-72.9. Subcommittee for Scleroderma Criteria of the American Rheuma-

tism Association Diagnostics and Therapeutic Criteria Committee:Preliminary criteria for the classification of systemic sclerosis(scleroderma). Arthritis Rheum 1980, 21:581-590.

10. Medsger TA Jr, Silman AJ, Steen VD, Black CM, Akesson A, BaconPA, Harris CA, Jablonska S, Jayson MI, Jimenez SA, et al.: A dis-ease severity scale for systemic sclerosis: development andtesting. J Rheumatol 1999, 26:2159-2167.

11. Kazerooni EA, Martinez FJ, Flint A, Jamadar DA, Gross BH, Spiz-arny DL, Cascade PN, Whyte RI, Lynch JP 3rd, Toews G: Thin-section CT obtained at 10-mm increments versus limitedthree-level thin-section CT for idiopathic pulmonary fibrosis:correlation with pathologic scoring. AJR Am J Roentgenol1997, 169:977-983.

12. Meloni F, Caporali R, Marone Bianco A, Paschetto E, Morosini M,Fietta AM, Patrizio V, Bobbio-Pallavicini F, Pozzi E, Montecucco C:BAL cytokine profile in different interstitial lung diseases: afocus on systemic sclerosis. Sarcoidosis Vasc Diffuse Lung Dis2004, 21:111-118.

13. Report of the European Society of Pneumology Task Group: Tech-nical recommendations and guidelines for bronchoalveolarlavage (BAL). Eur Respir J 1989, 2:561-585.

14. Görg A, Postel W, Günther S: The current state of two-dimen-sional electrophoresis with immobilized pH gradients. Electro-phoresis 1988, 9:531-546.

15. Candiano G, Bruschi M, Musante L, Santucci L, Ghiggeri GM,Carnemolla B, Orecchia P, Zardi L, Righetti PG: Blue Silver: avery sensitive colloidal Coomassie G-250 staining for pro-teome analysis. Electrophoresis 2004, 25:1327-1333.

16. Wilkins MR, Appel RD, Van Eyk JE, Chung MCM, Gorg A, HeckerM, Huber LA, Langen H, Link AJ, Paik Y, et al.: Guidelines for thenext 10 years of proteomics. Proteomics 2006, 6:4-8.

17. ExPASY Proteomics Server [http://www.expasy.ch/]18. Ludwicka-Bradley A, Bogatkevich G, Silver RM: Thrombin-medi-

ated cellular events in pulmonary fibrosis associated with sys-temic sclerosis (scleroderma). Clin Exp Rheumatol 2004, 2(3Suppl 33):S38-S46.

19. Kahaleh MB: Vascular involvement in systemic sclerosis (SSc).Clin Exp Rheumatol 2004, 2(3 Suppl 33):S19-S23.

20. Steen VD: The lung in systemic sclerosis. J Clin Rheumatol2005, 11:40-46.

21. Pannu J, Trojanowska M: Recent advances in fibroblast signal-ing and biology in scleroderma. Curr Opin Rheumatol 2004,16:739-745.

22. Postlethwaite AE, Shigemitsu H, Kanangat S: Cellular origins offibroblasts: possible implications for organ fibrosis in sys-temic sclerosis. Curr Opin Rheumatol 2004, 16:733-738.

23. Christner PJ, Jimenez SA: Animal models of systemic sclerosis:insights into systemic sclerosis pathogenesis and potentialtherapeutic approaches. Curr Opin Rheumatol 2004,16:746-752.

24. Atamas SP, White B: The role of chemokines in the pathogen-esis of scleroderma. Curr Opin Rheumatol 2003, 15:772-777.

25. Kinnula VL, Fattman CL, Tan RJ, Oury TD: Oxidative stress in pul-monary fibrosis: a possible role for redox modulatory therapy.Am J Respir Crit Care Med 2005, 172:417-422.

26. Kinnula VL: Focus on antioxidant enzymes and antioxidantstrategies smoking in related airway diseases. Thorax 2005,60:693-700.











































http://www.expasy.ch/

















27. Comhair SA, Erzurum SC: Antioxidant responses to oxidant-mediated lung diseases. Am J Physiol Lung Cell Mol Physiol2002, 283:L246-L255.

28. Hayes JD, Flanagan JU, Jowsey IR: Glutathione transferases.Annu Rev Pharmacol Toxicol 2005, 45:51-88.

29. Kinnula VL, Crapo JD: Superoxide dismutases in the lung andhuman lung diseases. Am J Respir Crit Care Med 2003,167:1600-1619.

30. Tourkina E, Gooz P, Oates JC, Ludwicka-Bradley A, Silver RM,Hoffman S: Curcumin-induced apoptosis in scleroderma lungfibroblasts: role of protein kinase cepsilon. Am J Respir CellMol Biol 2004, 31:28-35.

31. Abrahamson M, Alvarez-Fernandez M, Nathanson CM: Cystatins.Biochem Soc Symp 2003, 70:179-199.

32. Dickinson DP: Cysteine peptidases of mammals: their biologi-cal roles and potential effects in the oral cavity and other tis-sues in health and disease. Crit Rev Oral Biol Med 2002,13:238-275.

33. Kato T, Imatani T, Miura T, Minaguchi K, Saitoh E, Okuda K:Cytokine-inducing activity of family 2 cystatins. Biol Chem2000, 381:1143-1147.

34. Kato T, Ito T, Imatani T, Minaguchi K, Saitoh E, Okuda K: CystatinSA, a cysteine proteinase inhibitor, induces interferon-γexpression in CD4-positive T cells. Biol Chem 2004,385:419-422.

35. Zavasnik-Bergant T, Repnik U, Schweiger A, Romih R, Jeras M,Turk V, Kos J: Differentiation- and maturation-dependent con-tent, localization, and secretion of cystatin C in human den-dritic cells. J Leukoc Biol 2005, 78:122-134.

36. Gabay C, Kushner I: Acute-phase proteins and other systemicresponses to inflammation. N Engl J Med 1999, 340:448-454.

37. Honore B, Vorum H: The CREC family, a novel family of multipleEF-hand, low-affinity Ca2+-binding proteins localised to thesecretory pathway of mammalian cells. FEBS Lett 2000,466:11-18.

38. Vorum H, Hager H, Christensen BM, Nielsen S, Honore B: Humancalumenin localizes to the secretory pathway and is secretedto the medium. Exp Cell Res 1999, 248:473-481.

39. Basrur V, Yang F, Kushimoto T, Higashimoto Y, Yasumoto K,Valencia J, Muller J, Vieira WD, Watabe H, Shabanowitz J, et al.:Proteomic analysis of early melanosomes: identification ofnovel melanosomal proteins. J Proteome Res 2003, 2:69-79.

40. Vorum H, Jacobsen C, Honore B: Calumenin interacts withserum amyloid P component. FEBS Lett 2000, 465:129-134.

41. Coppinger JA, Cagney G, Toomey S, Kislinger T, Belton O,McRedmond JP, Cahill DJ, Emili A, Fitzgerald DJ, Maguire PB:Characterization of the proteins released from activated plate-lets leads to localization of novel platelet proteins in humanatherosclerotic lesions. Blood 2004, 103:2096-2104.

42. Chardin P, Paris S, Antonny B, Robineau S, Beraud-Dufour S,Jackson CL, Chabre M: A human exchange factor for ARF con-tains Sec7- and pleckstrin-homology domains. Nature 1996,384:481-484.

43. Venkateswarlu K: Interaction protein for cytohesin exchangefactors 1 (IPCEF1) binds cytohesin 2 and modifies its activity.J Biol Chem 2003, 278:43460-43469.

44. Theis MG, Knorre A, Kellersch B, Moelleken J, Wieland F, KolanusW, Famulok M: Discriminatory aptamer reveals serumresponse element transcription regulated by cytohesin-2.Proc Natl Acad Sci USA 2004, 101:11221-11226.

45. Ryckman C, Vandal K, Rouleau P, Talbot M, Tessier PA: Proin-flammatory activities of S100: proteins S100A8, S100A9, andS100A8/A9 induce neutrophil chemotaxis and adhesion. JImmunol 2003, 170:3233-3242.

46. Ahmad A, Bayley DL, He S, Stockley RA: Myeloid related protein-8/14 stimulates interleukin-8 production in airway epithelialcells. Am J Respir Cell Mol Biol 2003, 29:523-530.

47. Viemann D, Strey A, Janning A, Jurk K, Klimmek K, Vogl T, HironoK, Ichida F, Foell D, Kehrel B, et al.: Myeloid-related proteins 8and 14 induce a specific inflammatory response in humanmicrovascular endothelial cells. Blood 2005, 105:2955-2962.

48. Vogl T, Ludwig S, Goebeler M, Strey A, Thorey IS, Reichelt R, FoellD, Gerke V, Manitz MP, Nacken W, et al.: MRP8 and MRP14 con-trol microtubule reorganization during transendothelial migra-tion of phagocytes. Blood 2004, 104:4260-4268.

49. Shibata F, Miyama K, Shinoda F, Mizumoto J, Takano K, NakagawaH: Fibroblast growth-stimulating activity of S100A9 (MRP-14).Eur J Biochem 2004, 271:2137-2143.

50. Zhang H, Meng LH, Zimonjic DB, Popescu NC, Pommier Y: Thir-teen-exon-motif segnature for vertebrate nuclear and mito-chondrial type IB topoisomerases. Nucleic Acids Res 2004,32:2087-2092.

51. Steen VD, Powell DL, Medsger TA Jr: Clinical correlations andprognosis based on serum autoantibodies in patients withsystemic sclerosis. Arthritis Rheum 1988, 31:196-203.




























































Documents

Analysis of bronchoalveolar lavage fluid proteome from systemic sclerosis patients with or without functional, clinical and radiological signs of lung fibrosis